AC 2012-3976: POWER AND EFFICIENCY MEASUREMENT IN A THER-MOELECTRIC GENERATOR

Dr. Steven O’Halloran, University of Portland

Steven O’Halloran is an Assistant Professor of mechanical engineering at the University of Portland (UP)and teaches courses in the thermal/fluid sciences. He conducts research in the area of heat exchangersincluding both experimental work and numerical simulations. O’Halloran received his B.S. (2000), M.S.(2002), and Ph.D. (2006) degrees in mechanical engineering from Kansas State University.

Mr. Matthew Rodrigues, University of Portland

Matthew Rodrigues is a senior in mechanical engineering at the University of Portland, Donald P. ShileySchool of Engineering.

c©American Society for Engineering Education, 2012

Power and Efficiency Measurement in a Thermoelectric Generator

Abstract

This paper documents a laboratory experiment to determine the efficiency of a commercial

thermoelectric generator. Thermoelectric devices make use of the Seebeck effect to convert a

difference in temperature into electricity. Thermoelectric devices are used or proposed for a

number of different current applications including waste heat recovery. In the experiment, a

LabVIEW data acquisition system measures real-time efficiency of the thermoelectric device

under different testing conditions. The experiment is well-suited for an undergraduate thermal

sciences lab or a demonstration for a thermodynamics lecture course. The experiment is also

relatively low-cost aside from the data acquisition system. Results from the system as well as a

detailed setup are presented.

Introduction

Thermoelectric generators convert heat into electricity through the Seebeck effect. The devices

have no moving parts and can have a relatively long life which is a benefit. However,

thermoelectric generators have lower efficiency when compared to many other power generation

devices. Thermoelectric generators can be used in engineering applications ranging from

automotive waste heat recovery1 to powering a small wristwatch

2.

Thermoelectric devices have been incorporated into undergraduate courses in several different

ways3,4,5,6

. The devices offer the advantage of relatively low cost and small size, while still

providing an excellent thermodynamic example of power generation. Although common

undergraduate thermodynamic textbooks mention thermoelectric devices7,8

, as of now the

content related to thermoelectric devices is fairly limited. Therefore, laboratory use of

thermoelectric devices can help provide further information on the topic in an undergraduate

setting.

The application of interest in this paper, and what sets this work apart from the previous work on

thermoelectric devices, is the instantaneous measurement of efficiency for a commercial

thermoelectric generator using LabVIEW data acquisition. Thermal efficiency is an important

parameter and provides important information for designing systems. The setup described can

be incorporated into an undergraduate thermodynamics laboratory, or also used as a

demonstration in a thermodynamics lecture course.

Experimental Setup

A commercial thermoelectric device was used for the experimental testing and is a model TEC1-

12706 Bismuth Telluride device with a physical size of 40mm x 40mm x 3.5 mm. The device

has 127 couples and a photo of the device is shown in Figure 1 below.

Figure 1: Photo of thermoelectric device used for testing (model TEC1-12706).

A testing assembly was constructed such that a known heat could be added to “hot” side of the

device. By measuring the power output of the thermoelectric device through a load, the

efficiency of the thermoelectric device can be calculated as follows:

(1)

Where, η = thermal efficiency

Pout = measured power output of the device (watts)

Qin = measured input heat to the device (watts)

An assembly was created to house the thermoelectric device for testing purposes. A diagram of

the assembly is shown in Figure 2 with both an assembled view and exploded view. Starting at

the bottom of the device, an end plate fabricated from copper is used to attach a thin Omega

Kapton heater to the bottom plate which is also made from copper. Copper was chosen for

construction due to its high thermal conductivity so that each plate will distribute heat well. The

bottom plate has a small hole drilled in the side to accommodate a type-T thermocouple for

temperature measurement. Above the bottom plate is the thermoelectric device (pictured in

Figure 1). Silver conductive thermal grease was used on both sides of the thermoelectric device

to provide good thermal contact in the assembly. Above the thermoelectric device is the top

plate which is fabricated from copper and contains a channel for water cooling. The entire

assembly is held together with plastic screws from the top plate to the end plate. Plastic was

chosen as the material for the screws to reduce heat transfer through the screws. The assembly is

also surrounded with rigid insulation (2” thick housing type insulation) on all sides except for the

top plate to ensure that the heat provided by the heater is transferred through the thermoelectric

device and not dissipated elsewhere. The insulation is not shown on Figure 2, but it can be seen

later in Figure 4.

Figure 2: (a) assembled view of thermoelectric testing device with dimensions in millimeters and (b) exploded view of

thermoelectric testing device.

A block diagram of the overall experimental setup is shown in Figure 3. In addition to the

thermoelectric testing device described before, several other items are included. A data

acquisition system and laptop are used for recording data. A small DC pump and cooling water

tank (one gallon water container) are used to provide cooling for the system. A power supply is

used to provide power both to the heater of the thermoelectric testing device and also the pump

(a single power supply provides power independently to the two devices). Two resistors, 0.488

Ω and a 0-25 Ω rheostat, are used to provide the load for testing. A variable resistance was

selected to best match the load resistance to the device resistance to provide maximum power. It

also allows the user to easily see the effect of load resistance on power output, voltage output,

and efficiency. A photograph of the actual test setup is shown in Figure 4 and a detailed listing

of all of the equipment used including equipment cost is shown in Table 1.

Figure 3: Schematic of overall experimental setup.

Figure 4: Photo of overall experimental setup.

Thermoelectric testing device

Cooling water tank

Pump

Power supply

Load

Laptop

DAQ

0.488 Ωresistor

0-25 Ωrheostat

Cooling water

Table 1: List of Equipment Used

Item Description Cost

Thermoelectric device Model TEC1-12706 $3

Laptop Computer Lenovo T400 $1000

DAQ National Instruments cDAQ-9172 $1000

DAQ Analog Input Module NI 9215 $400

DAQ Temperature Input Module NI 9211 $300

Heater Omega Kapton flexible heater, 0.5” x 2”,

nominal resistance: 8.8 Ω (part # KH-

KIT-EFH-15001)

$30

DC Power Supply GW Instek GPC-3030D $400

Pump Cole Parmer mini DC gear pump (part #

07012-20) $59

Load Resistors Radio Shack 0.488 Ω (measured), 5 watt

resistor (part #271-0130) and Radio Shack

0-25 Ω rheostat, 3 watt (part #271-0265)

in series.

$7

A LabVIEW program was created to measure the performance of the thermoelectric device

(including instantaneous efficiency calculation). Five analog input channels were used to

measure information from the setup and those inputs are listed below.

Temperature of the hot side of the thermoelectric device (Type-T thermocouple)

Temperature of the cold side of the thermoelectric device (Type-T thermocouple)

Heater voltage, measured using a voltage divider (analog voltage input)

Load voltage, measured across variable resistor

Partial load voltage, measured across the 0.488Ω fixed resistor and used to calculate load

current.

In order to calculate efficiency using Equation (1), the input heat, Qin, and output power Pout,

must be found. The input heat is found using the heater voltage and heater resistance as shown

in Equation (2) below.

(2)

The heater voltage was measured by the data acquisition system and the heater resistance was

measured as 9.00Ω prior to testing.

The power output was calculated using the load voltage and load current with equation (3)

below.

(3)

The load voltage (Vload) is measured directly with the data acquisition system by measuring the

voltage drop across both the 0.488Ω fixed resistor and the 0-25Ω rheostat. The load current

(Iload) is found by measuring the voltage drop across only the 0.488Ω fixed resistor and

calculating the current.

A screenshot of the block diagram of the LabVIEW program is shown in Figure 5 and a

screenshot of the front panel is shown in Figure 6. The front panel allows the user to monitor the

hot and cold side temperatures of the thermoelectric device (graph on the left side of Figure 6) as

well as monitor the instantaneous efficiency of the device (graph on the right side of Figure 6).

It is very easy to visually see the effect that changing the load resistance has on the device

efficiency as seen in the spike in efficiency in Figure 6 (the load resistance was changed during

testing at this spike).

Figure 5: Block diagram of LabVIEW program.

Figure 6: Front panel of LabVIEW program.

Results

Six different tests were conducted with the commercial thermoelectric device. The heater

voltage was tested at 12, 14, 16, 18, 20, and 22 volts. Increasing the heater voltage increased the

temperature difference across the device. The key parameters of interest were the power output

and efficiency of the thermoelectric device for a range of temperature differences and load

voltages. The power output of the device depends on the output voltage (controlled by the load

resistance), and this relationship changes for various temperature differences. Figure 7 shows the

relationship between load power and voltage for each of the six tests. The data sets for each test

were created by varying the load resistance using the rheostat. The lowest temperature

difference tested was 20.6°C (corresponding to a heater voltage of 12 volts) and the highest

temperature difference tested was 68.1°C (corresponding to a heater voltage of 22 volts). As

expected the highest temperature difference produced the highest power output. For each curve,

the highest power output occurs at the point at which the load resistance matches the device

resistance. This allows the user to quickly see the benefits of load matching.

Figure 7 – Power output of thermoelectric device.

The efficiency of the device throughout the test was also calculated. Figure 8, below, shows the

relationship between efficiency and load voltage for each of the six tests. The efficiency

increases with increasing change in temperature and with the highest temperature change tested

(68.1°C), the maximum efficiency was 2.22%.

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Loa

d P

ow

er (W

atts

)

Load Voltage (Volts)

ΔT=20.6°C

ΔT=26.6°C

ΔT=35.3°C

ΔT=44.9°C

ΔT=55.9°C

ΔT=68.1°C

Figure 8 – Efficiency of thermoelectric device.

Table 2 shows data for each test including input heater voltage, average change in temperature,

maximum efficiency, maximum power generated, calculated load resistance at maximum

power/efficiency, and load voltage at maximum power/efficiency. The data collected can be

used to predict the performance of the commercial devices in other applications.

0.0

0.5

1.0

1.5

2.0

2.5

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Effi

cie

ncy

(%

)

Load Voltage (Volts)

ΔT=20.6°C

ΔT=26.6°CΔT=35.3°C

ΔT=44.9°CΔT=55.9°C

ΔT=68.1°C

Table 2 - Values at Maximum Efficiency

Input Heater

Voltage (V) Average ΔT (°C )

Maximum Efficiency

(%)

Maximum Power

Generated (W)

Calculated Load Resistance (Ω)

(at max efficiency/power)

Load Voltage (V) (at max

efficiency/power)

12 20.6 0.71 0.11 3.01 0.58

14 26.6 1.03 0.23 2.81 0.80

16 35.3 1.26 0.36 3.02 1.04

18 44.9 1.52 0.55 2.76 1.23

20 55.9 1.82 0.81 3.38 1.66

22 68.1 2.22 1.17 2.81 1.82

Summary and Conclusions

This paper describes a LabVIEW based experimental setup for measuring the instantaneous

power generation and efficiency of a commercial thermoelectric device. Results are presented

for a particular Bismuth Telluride thermoelectric device (TEC1-12706). The load resistance is

variable in the experimental setup and the power generation and efficiency are both plotted

versus the voltage produced. The maximum temperature difference tested was 68.1°C and this

produced an efficiency of 2.22% and an output power of 1.17 watts. While this efficiency might

seem low, thermoelectric generators are noted for their relatively low conversion efficiency.

Also, the maximum temperature difference tested (68.1°C) is fairly modest, higher temperature

differences would result in higher efficiency. Typical thermoelectric devices require a

temperature difference of approximately 500°C to achieve an efficiency of 10% 9,10

. Testing is

planned for the future with higher temperature differences, but the current testing has been

successful for the conditions tested.

It is planned to use this experimental setup both as an in-class demonstration for a lecture-based

thermodynamics course as well as in a thermodynamics laboratory course. Aside from the data

acquisition system (a National Instruments LabVIEW based system in this case) and laptop

computer, the cost for the experimental setup is relatively modest and many of the needed

components are likely already available in a thermodynamics laboratory. This lab experiment

would have the advantage of exposing students to thermoelectric devices as well as incorporating

data acquisition for real-time efficiency calculations.

Acknowledgements

The authors would like to thank Alex Brotherston, (graduate of the University of Portland in

May 2011, electrical engineering) for his help in development of the experimental setup and

LabVIEW program.

Bibliography

1. Zorbas, K.T., E. Hatzikraniotis, and K.M. Paraskevopoulos. “Power and Efficiency

Calculation in Commercial TEG and Application in Wasted Heat Recovery in Automobile,”

Proceedings of the 5th European Conference on Thermoelectrics, September 10-12, 2007,

Odessa, Ukraine.

2. Leonov, V. and R.J.M. Vullers. “Thermoelectric Generators on Living Beings,” Proceedings

of the 5th

European Conference on Thermoelectrics, September 10-12, 2007, Odessa,

Ukraine.

3. Rutar, T. and G. Mason. “Design of Experiments in Introduction to Thermodynamics

Course,” 2011 ASEE Annual Conference and Exhibition, Vancouver, Canada.

4. Pantchenko, O., D.S. Tate, D. O’Leary, M.S. Isaacson, and A. Shakouri. “Enhancing Student

Learning Through Hands-On Laboratory Experiments on Renewable Energy Sources,” 2011

ASEE Annual Conference and Exhibition, Vancouver, Canada.

5. Glasgow, L. and D. Soldan. “Reconnecting Chemical Engineering Students with the Physical

World,” 2010 ASEE Annual Conference and Exhibition, Louisville, Kentucky.

6. Rosen, W. and E. Carr. “An Application-Based Approach to Introducing Microcontrollers to

First-Year Engineering Students,” 2010 ASEE Annual Conference and Exhibition,

Louisville, Kentucky.

7. Moran, M.J. and Shapiro, H.N., Fundamentals of Engineering Thermodynamics, 6th Edition,

Wiley, 2008, p. 549.

8. Cengel, Y.A. and Boles, M.A., Thermodynamics: An Engineering Approach, 6th Edition,

McGraw-Hill, 2008, p. 651.

9. Rowe, D.M. “Thermoelectric Waste Heat Recovery as a Renewable Energy Source,”

International Journal of Innovations in Energy Systems and Power, Vol.1, no.1. November,

2006.

10. Snyder, G. Jeffrey. “Small Thermoelectric Generators,” The Electrochemical Society

Interface, Fall 2008.